The Curate’s Egg: A Critical Analysis of Michael Liebreich’s Deep Dive into Hydrogen [GasTransitions]

What follows is a critique of a two-part article on hydrogen written by Michael Liebreich, a senior contributor with BNEF and member of the UK’s Board of Trade. The articles were published on the BNEF site under its Blog section. The first part of my critique covers generalities, the second part is in the style of a fact checker.

Since Liebreich is an influential energy thinker, it is important to look closely at what he writes about hydrogen, one of the key energy technologies of today. Unfortunately, Liebreich does not seem to be able to come to terms with his subject. He mixes praise with criticism and in the end leaves the reader guessing as to what he really thinks of the potential of hydrogen. Although the articles make some good points, these are outweighed by many errors and contradictions. 

For example, in Part 1 the 3^rd paragraph Liebreich notes that the storage of hydrogen “requires compression to 700 times atmospheric pressure, refrigeration to minus 253 degrees Celsius or combining with an organic chemical or metal hydride”.  

This ignores the fact that an industrial hydrogen network spanning the Benelux and France seems to do quite well with 90bar and no liquefaction. 90bar is typical of the operating pressures for many gas networks including natural gas. However, the blog then says (in Part 2 under “Batch Processing”) “Hydrogen could have the advantage here, since it can be delivered by pipeline at very high rates, or stored on site, in compressed gas or liquid form” So, first there is a problem transporting hydrogen, then things are not so bad?

Later in Part 2 the writer contradicts the earlier positive comments: “hydrogen pipelines cost three times as much as power lines, and ships and trucks are even worse”. Apart from a few linking systems nobody in the EU is proposing new H₂ pipelines. What is proposed is that existing gas pipelines be re-purposed to carry H₂. As for the cost (3x electricity) this is plain wrong. The BritNed HVDC link and the BBL gas pipeline have roughly the same length (230 km) and cost roughly the same to build (€500million). The electricity cable can carry 8 TWh of electricity per year, the pipeline (derated for H₂) 40 TWh – 5x the amount of energy for the same price.

Another example of omitted context/data: (Part 1) “It (hydrogen) can embrittle metal; it escapes through the tiniest leaks; and, yes, it really is explosive”. It depends on what sort of metal. Embrittlement of steel due to hydrogen is a function of the type of steel used. Steel quality of X52 or less is unaffected by hydrogen. A report prepared for the UK government noted as much. Much of Europe’s high pressure and intermediate pressure gas networks are made of X52 steel. Until natural gas came along, most of the European gas network operated safely on coal gas, which was 50% hydrogen. Perhaps these inconvenient facts get in the way of a good story.

Factual omissions

The EU’s Green New Deal is reviewed, wrong facts are used and as a result, wrong conclusions drawn (see fact checker). The writer worries about where the money for all the spending on renewables and electrolysers will come from. This ignores the fact that renewables have a business case and can be funded through the usual mix of debt and equity.

Factual omissions reinforce the writer’s arguments, but when these omissions are included the arguments fall to pieces. (Part 1) “The working assumption for EU electrolyzer capacity in 2050,…is 500 GW. To put that in context, the maximum peak electrical load ever recorded for all of Europe was 546 GW”. The implication is that electrolysers would consume all EU electricity. The missing context is that to reach its 2030 targets the EU will need an installed base of PV and wind of around 900 GW and will need to keep building capacity rapidly after that date probably reaching something in the range 2,000-2,500 GW by 2050. Clearly, electrolysers taking 20% or maybe 25% of load rather than 95% does not make such a good story.

Questions are posed by the writer about hydrogen pricing but never really answered. The fact checker provides the answers that BNEF was unwilling or unable to supply.

Contradictions: Electrolysers and renewables are becoming cornerstones of EU industrial policy. However, according to the writer this reality will not matter because cheap Chinese electrolysers will win, even though it is recognised by the writer that at higher capacity factors CAPEX does not matter.

A discussion on the use of surplus renewable electricity to make hydrogen includes: “… very idea of using surplus renewable energy to generate hydrogen will turn out to be…. a mirage”. This is based on the assertion that “the only thing that matters is to produce the cheapest green hydrogen possible, or you will be outcompeted by producers using the lowest-cost renewable electricity at high capacity factors, delivering via pipeline”. But as the writer observed at the beginning of the article (paraphrasing) “hydrogen is difficult to transport”. A straw man is then built to justify the argument: “Imagine, for the sake of argument, a future grid with such huge penetration of variable renewable generation that curtailment reaches 33%.”.

Surplus renewable electricity is the trajectory on which the EU is embarked. Renewables will by 2030 to 2035 account for around 80% of the electricity mix. For renewables to account for 80% they will need to produce a surplus because once renewables account for more than around 40% of demand, there is always some renewable electricity which cannot be used. To meet 80% of demand, renewables need to produce 120% of demand and this of course means that there is a surplus of 40% (the difference between 80% and 120%). The only scalable technology able to absorb such quantities of electricity will be electrolysers.

Wrong costs

Part 2 includes a price comparison (which is factually incorrect – see fact checker) between green H₂ and blue H₂ but omits consideration of the CO₂ footprint. In the case of green this is around 8gCO₂/kWh. (A letter in Nature Energy of 2017 noted that the carbon footprint of a kWh from PV and wind systems was 6gCO₂/kWh and 4gCO₂/kWh respectively.)

By contrast, H₂ from a steam methane reforming (SMR)+CCS system will have a footprint of at least 25gCO₂/kWh as CCS captures at best (yet to be proven) 90% of the CO₂. Furthermore, this does not include methane leakage – an area of increasing interest to the Commission. Thus, we are a very long way away from “zero-carbon hydrogen produced via fossil fuels with carbon capture”.

In Part 2 covering steel, out-of-date BNEF data is used to make an out-of-date statement: “hydrogen-based steel would become competitive with the most expensive current steel production as soon as it can be made for 2.5 euros per kg, which is any time now”. As the fact checker shows H₂ can be made for €1.65/kg right now. The article also uses the same incorrect facts to exaggerate a situation. In the case of glass, the wrong costs for green H₂ are used to imply that a combination of green H₂ and ETS could never bridge the price gap with natural gas.

Remarkably, given BNEF’s knowledge of the renewables sector and price developments, this statement is made: “By 2050, green hydrogen may achieve a price of $0.80/kg, dependent on directly connected renewable power being available at $14 to $17/MWh”. This suggests that BNEF is unaware of recent PV auction results in Portugal (see fact checker).

In Part 2, there is a section on water and space heating in which the writer demonstrates a profound lack of knowledge. In summary, modelling done by PWR and Challoch Energy has analysed local power networks (I am an engineer specialised in distribution networks). These networks will run out of capacity long before heat pumps reach even 40% penetration. Electric vehicle charging makes things worse. H₂-powerred fuel cells or other micro-CHP systems delivering heat and electricity provide embedded network support. They facilitate more heat pumps and more EVs. Reinforcing underground distribution networks to support heat pumps and EVs is costly, disruptive and very time consuming.

The final substantive section covering “power system” includes naive assertions on renewables and their penetration into power systems. As renewables grow in any electrical system, so there is a parallel growth in surplus renewable electricity that no amount of demand response or batteries can halt. As already noted, if renewables for a period of months account for 80% of electricity demand then actual renewable electricity production will be 120% of demand. Which means that 40% of the 120% is looking for a home. Demand response and batteries are functionally incapable of providing that home. Neither will cross-border connections (mentioned as part of the solution). The reasons why this is so are complex and require more space than this blog permits.

The overall impression is of a pair of documents written by a Mr Hyde who hates green hydrogen and a Dr Jekyll who thinks that green hydrogen has a bright future.

The final “Coda” suggests that BNEF is out of touch with developments on the ground. Taking one example: “What this means is that we should forget “hydrogen homesteads” – homes and small communities trying to get off-grid using hydrogen”. There are several energy communities in Germany that connected renewables to electrolysers and use the hydrogen locally. I have spoken to people in one and they know of others.

Mike Parr is an independent technical energy consultant based in Brussels. He is co-founder, with Simon Minett, of Green H2 for Europe (GH24EU), focused on promoting the fast development of hydrogen produced only from renewable electricity.